Re: Alternative to ListContourPlot3D
- To: mathgroup at smc.vnet.net
- Subject: [mg40556] Re: Alternative to ListContourPlot3D
- From: Jens-Peer Kuska <kuska at informatik.uni-leipzig.de>
- Date: Thu, 10 Apr 2003 03:37:01 -0400 (EDT)
- Organization: Universitaet Leipzig
- References: <b70o7a$9kh$1@smc.vnet.net>
- Reply-to: kuska at informatik.uni-leipzig.de
- Sender: owner-wri-mathgroup at wolfram.com
Hi, http://phong.informatik.uni-leipzig.de/~kuska/mathgl3dv3/ with it's MVListContourPlot3D[] function ? it find and render 50 000 polygons per second on a modern Intel/AMD CPU with OpenGL. The tutorial has several examples with medical data from CT scans. You man still need a huge memory -- a 512 x 512 x 200 voxesl data set of a human heart generate 1.7 million polygons and need 1.2 GByte RAM. For these high resolved data sets MathGL3d has a ListDensityPlot3D[] that make true OpenGL based volume rendering with a few hundred polygons. The 3.1 beta version support shading for the (List)DensityPlot3D[] function on nVidia cards better with GeForce 3 chips or better Regards Jens "Y.A.Tesiram" wrote: > > Dear Group, > ListContourPlot3D can be excruciating slow when trying to construct a 3D > image out of a series of (say 40 to 50) high resolution 2D (512 * 512 > matrix) images. Can someone suggest an alternate strategy for handling > these types of data sets. > As a short example, take 10 4x4 images. > > testdata = Range[10*4*4]; > twoDimages = Partition[Partition[testdata,4],4]; > > You can use Raster or ListDensityPlot or ListContourPlot to look at each > image. > > e.g. ListDensityPlot[twoDimages[[1]], Mesh->False, PlotRange->All] > > will give you a 2D image of the first image in that set of which there are > 10. > > I want to put all these images together as a 3D volume. ListContourPlot is > what I have found and with large data sets it is really really slow. What > are my options? > > Thanks > Yas